Navigating Prompt Complexity for Zero-Shot Classification: A Study of Large Language Models in Computational Social Science